metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: start
dtype: float64
- name: end
dtype: float64
- name: transcription
dtype: string
- name: eaf_source
dtype: string
- name: wav_source
dtype: string
- name: raw_transcription
dtype: string
- name: clip
dtype: string
- name: wav_rawpath
dtype: string
- name: path
dtype: string
- name: allosaurus
dtype: string
- name: clap_ipa_cos_sim
dtype: float64
- name: wada_snr
dtype: float64
- name: nist_stnr
dtype: float64
- name: speaker-diarization-3.1
dtype: string
- name: voice-activity-detection
dtype: string
- name: whisper-large-v3
dtype: string
- name: clapipa-transcription-allosaurus
dtype: float64
- name: vad_s
dtype: float64
- name: drz_s
dtype: float64
- name: trans_len
dtype: int64
- name: pcnt_speech
dtype: float64
- name: trans_len_sq
dtype: float64
- name: trans_len_log
dtype: float64
- name: cos_sim_softmax
dtype: float64
- name: cos_sim_log
dtype: float64
- name: duration
dtype: float64
splits:
- name: train
num_bytes: 1106517483.8
num_examples: 16384
- name: validation
num_bytes: 131592822.088
num_examples: 2048
- name: test
num_bytes: 76685401.888
num_examples: 2048
download_size: 1290401128
dataset_size: 1314795707.776
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*